Skip to content

Conversation

trivialfis
Copy link
Member

@trivialfis trivialfis commented Oct 4, 2025

Minor optimization for building deep trees.

todo: test extmem.

@trivialfis trivialfis requested a review from Copilot October 4, 2025 00:56
Copy link
Contributor

@Copilot Copilot AI left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull Request Overview

This PR optimizes histogram building for deep trees by using GPU streams instead of directly passing CUDA contexts. The change allows for better concurrency and resource utilization during histogram construction.

Key changes:

  • Modified histogram building methods to accept stream references instead of CUDA contexts
  • Added stream pool for managing multiple concurrent histogram builds
  • Updated all test cases to use the new stream-based API

Reviewed Changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 2 comments.

File Description
tests/cpp/tree/gpu_hist/test_histogram.cu Updated test calls to use stream references for histogram building
src/tree/updater_gpu_hist.cu Added stream pool and event synchronization for concurrent histogram builds
src/tree/gpu_hist/histogram.cuh Changed BuildHistogram method signature to accept stream reference
src/tree/gpu_hist/histogram.cu Updated histogram building implementation to use streams

Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant